Breast milk transmits human immunodeficiency virus (HIV), though only ~15% of infants breastfed by HIV-infected mothers become infected. Breastfed infants ingest ~105−108 maternal leukocytes daily, though these cells are understudied. Here we describe the isolation of breast milk leukocytes and an analysis of their phagocytic capacity.
Even in the absence of antiretroviral drugs, only ~15% of infants breastfed by HIV-infected mothers become infected, suggesting a strong protective effect of breast milk (BM). Unless access to clean water and appropriate infant formula is reliable, the WHO does not recommend cessation of breastfeeding for HIV-infected mothers. Numerous factors likely work in tandem to reduce BM transmission. Breastfed infants ingest ~105−108 maternal leukocytes daily, though what remains largely unclear is the contribution of these cells to the antiviral qualities of BM. Presently we aimed to isolate cells from human BM in order to measure antibody-dependent cellular phagocytosis (ADCP), one of the most essential and pervasive innate immune responses, by BM phagocytes against HIV targets. Cells were isolated from 5 human BM samples obtained at various stages of lactation. Isolation was carried out via gentle centrifugation followed by careful removal of milk fat and repeated washing of the cell pellet. Fluorescent beads coated with HIV envelope (Env) epitope were used as targets for analysis of ADCP. Cells were stained with the CD45 surface marker to identify leukocytes. It was found that ADCP activity was significant above control experiments and reproducibly measurable using an HIV-specific antibody 830A.
Human breast milk (BM) is comprised of maternal cells that are >90% viable1. Cell composition is impacted strongly by stage of lactation, health status of mother and infant, and individual variation, which remains poorly understood1,2,3,4. Given that BM contains ~103−105 leukocytes/mL, it can be estimated that breastfed infants ingest ~105−108 maternal leukocytes daily5. Various in vivo studies have demonstrated that maternal leukocytes provide critical immunity to the infant and are functional well beyond these sites of initial ingestion5,6,7,8,9,10,11. All maternally-derived cells ingested by the infant have the potential to perform immune functions alongside or to compensate for the infant's own leukocytes12.
Mother-to-child transmission (MTCT) of human immunodeficiency virus (HIV) remains a crisis in resource-limited countries. As diarrheal and respiratory diseases are responsible for substantial rates of mortality among infants in resource-limited countries, and these illnesses are significantly reduced by exclusive breastfeeding, the benefits to HIV-infected mothers of breastfeeding far outweigh the risks13,14,15. Unless access to clean water and appropriate infant formula is reliable, the WHO does not recommend cessation of breastfeeding for HIV-infected mothers16. Approximately 100,000 MTCTs via BM occur annually; yet, only ~15% of infants breastfed by their HIV-infected mothers become infected, suggesting a strong protective effect of BM17,18,19,20,21. Numerous factors likely work in tandem to prevent transmission. Importantly, HIV-specific antibodies (Abs) in BM have been correlated with reduced MTCT and/or reduced infant death from HIV infection22,23. What remains largely unclear is the contribution of the cellular fraction of BM to its antiviral qualities.
Many Abs facilitate a variety of anti-viral activities mediated by the 'constant' region of the immunoglobulin molecule, the crystallizable fragment (Fc), via interaction with Fc receptors (FcRs) found on virtually all innate immune cells, virtually all of which are found in human BM24. Antibody-dependent cellular phagocytosis (ADCP) has been demonstrated as necessary for the clearance of viral infections and has been understudied in the case of prevention of MTCT of HIV25,26,27,28,29. Given the paucity of knowledge about the potential contribution of ADCP activity by BM phagocytes to prevention of MTCT of HIV, we aimed to develop a rigorous method to isolate cells from human BM in order to undertake a study of ADCP mediated by cells from BM obtained at various stages of lactation.
Each participant in this study was recruited and interviewed in accord with the ethical and institutional review board (IRB) approval with the guidance and authorization of Mount Sinai’s Program for the Protection of Human Subjects (PPHS) using an IRB-approved protocol for obtaining breast milk samples.
1. Target Microsphere preparation
2. Antibody-Dependent Cellular Phagocytosis (ADCP) Assay Plate Preparation
3. Breast milk cell isolation
4. ADCP assay
NOTE: Methods described here are adapted from Ackerman et al.32.
5. Analysis by flow cytometry
Milk can be kept at room temperature or cooler (though not frozen); however, given that we have observed reduced viability when milk has been kept very cold (data not shown), and that it is simpler to collect, store briefly, and transport at ambient temperatures, it is recommended that samples are not refrigerated in order to reduce sample-to-sample variability. In milk obtained 7−183 days post-partum, cell concentration determined by automated cell counter ranged from 16,083−222,857 cells/mL. Figure 1 illustrates the gating strategy eliminating doublets, debris, and dead cells. Viability was ~90−99%. Approximately 1.6−12.3% of total live cells were CD45+ leukocytes (Table 1). Most purported monocytes appeared to be precursors/immature cells as previously described, based on the suggestive SSC vs. CD45 gating34. The purported monocytes were defined as SSClow-intermediate/CD45low, though few exhibited the higher CD45 staining levels distinct from the purported lymphocyte population (SSClow/CD45low) more typically associated with blood monocytes (Figure 1), similar to previous studies33,34. The purported granulocytes were defined as SSChigh/CD45intermediate33,34 (Table 1). Note that this classification is only suggestive and that lineage-specific markers would be needed to confirm cell type.
ADCP activity of the freshly isolated BM cells was measured using the HIV-specific human mAb 830A, which is specific for the V2 region of the HIV envelope and binds to the V1V2-2F5K antigen tested here. ADCP activity was measured for the example here using milk obtained at 1 month post-partum (Figure 2A). Example data shows the expected FITC (bead+) histograms seen when gating on CD45+ cells (data generated using 1 µg/mL mAb is shown). The black markers indicate the populations used to calculate ADCP scores. In the sample 830A data (first panel of Figure 2A), percentage of CD45+ cells and mean fluorescence of that population are shown, which were used to calculate the ADCP score using the equation in step 3.4. Cells pre-incubated with actin inhibitor cytochalasin-D (cytoD) and/or FcR-blocking Abs (FcBlock) prior to their incubation with the Ab-bound/antigen-coupled beads exhibited ADCP activity at the level of the control mAb 3865 or below, indicating ADCP was FcR and actin-dependent (Figure 2). The ADCP score determined for total CD45+ cells was ~25−35-fold above background levels defined using the negative control anti-anthrax mAb 3865. Each major subset was analyzed separately as well. The purported granulocytes exhibited ADCP activity ~12−29-fold higher than background. The purported monocyte ADCP was ~2−3-fold above background (Figure 2). The purported lymphocytes as expected did not exhibit any measurable ADCP activity (less than 3x standard deviation of the ADCP score AUC of the non-specific negative control mAb 3865; data not shown).
Figure 1: Sample flow cytometry data of cells isolated from breast milk. Cells were processed and stained as described in the protocol. (A) Single cells were gated on to eliminate doublets in an FSC-H vs. FSC-A plot as shown, also gating out the small debris <5,000 in FSC-A. (B) This population was used to gate on live cells (which do not stain with the viability dye) in an SSC vs. V450 (viability stain) plot. (C) These live cells were used in an FSC vs. SSC plot. The expected position of non-leukocytes, likely to be predominantly mammary epithelial cells, is highlighted ("E"). (D) The same FSC vs. SSC plot is shown only with CD45+ cells. The major leukocyte subsets noted are only purported identities based on well-established and expected SSC parameters (G = granulocytes; M = monocytes; L = lymphocytes). (E) Viable cells were used for an SSC vs. CD45 plot with the major leukocyte subsets noted. Back-gating from this plot yielded the data shown in panel D. Note that this classification is only suggestive and that lineage-specific markers are needed to confirm cell type. Please click here to view a larger version of this figure.
Figure 2: Sample ADCP data using cells isolated from breast milk. The ADCP assay performed is based on the assay adapted from Ackerman et al.30. The assay was performed as outlined in the protocol above. (A) Sample FITC histograms at 1 µg/mL mAb, with markers indicating the bead/FITC+ populations used to determine the ADCP score. Scores were calculated as (MFI of bead-positive cells) x (% of total CD45+ cells in the bead/FITC+ positive population). The first panel using mAb 830A alone also indicates the percentage of total CD45+ cells and the mean fluorescence intensity value used to calculate the ADCP score in that example. (B) ADCP scores at each mAb dilution assayed were used to calculate area-under-the-curve (AUC) values in graphics software. For control experiments, actin inhibitor cytochalasin-D (cytoD), FcR blocking agent (FcBlock), or a combination of both were pre-incubated with cells prior to their addition to the immune complexes (see legends). Note that this cell classification is only suggestive and that lineage-specific markers are needed to confirm cell type. Please click here to view a larger version of this figure.
Sample | Cells/mL | % CD45+ | % Granulocytes* | % Monocytes* |
1 | 222,857 | 12.3 ± 1.9 | 13.6 ± 3.8 | 65.9 ± 5.6 |
2 | 27,361 | 1.6 ± 0.01 | 25.2 ± 4.0 | 9.1 ± 5.6 |
3 | 161,486 | 3.6 ± 1.1 | 47.8 ± 6.8 | 24.3 ± 4.3 |
4 | 16,083 | 2.7 ± 0.1 | 17.9 ± 3.5 | 34.4 ± 1.0 |
5 | 25,155 | 4.0 ± 0.7 | 29.7 ± 2.6 | 20.5 ± 1.4 |
*of CD45+ cells |
Table 1: Examples of typical breast milk cell counts and characteristics.
The flow cytometry-based technique for measuring ADCP activity described herein was first described in 201130 and has since been proven robust and cited in more than 80 studies. The protocol described here adapts this technique for use with primary BM cells for the first time. Previous studies of Fc-mediated functionality by BM cells have been largely limited to measurement of oxidative bursts or histology-based phagocytosis assays using cells isolated from colostrum (0−4 days after birth). Virtually no studies have examined cells in human BM past the colostrum phase. Studies using colostral cells have generally concluded that the granulocytes in colostrum are less active than those isolated from blood, behaving as an 'exudate cell' that has moved into the extravascular space35, though conflicting studies have reported similar phagocytic and bactericidal capacities36.
For decades, traditional microscopy was used to identify BM leukocytes, and this type of visual identification may have led to cell misidentification1. Few studies have compared BM leukocyte composition beyond the first month of lactation, and most have focused on colostrum. The use of flow cytometry to identify cells is likely to accurately identify cells, though only a small number of BM studies have been done using this method, often with a very small sample number. Current studies have indicated that the leukocyte content of BM at all stages of lactation varies widely, ranging from ~104−7 x 105 leukocytes/mL in early colostrum, decreasing to 103−5 x 104 leukocytes/mL in mature milk, though all studies confirm that cell concentration and composition is impacted strongly by the stage of lactation1,2,3,4,5. As milk transitions to its mature composition, neutrophil concentration appears to increase, though such studies have not typically extended beyond the first month postpartum34.
The protocol described herein uses fluorescent beads as the phagocytic target, though it likely can be applied to study BM ADCP of a variety of more biologically relevant targets such as immune complexes and infected cells, triggered by various Ab isotypes and subclasses. A larger cell staining panel can be employed to further differentiate the leukocytes. Large studies will be essential to develop a comprehensive understanding of ADCP by these relevant primary cells. This protocol allows for the establishment of ADCP by BM leukocytes as a potential mechanism for reduction of MTCT of HIV, as well as other pathogens.
The authors have nothing to disclose.
We thank Dr. Susan Zolla-Pazner in the Department of Medicine and Department of Microbiology at the Icahn School of Medicine at Mount Sinai for manuscript review. The NIH/NICHD provided funding for this project under grant number R21 HD095772-01A1. In addition, R. Powell was supported by funds from the Department of Medicine, Division of Infectious Diseases, Icahn School of Medicine at Mount Sinai.
1 µm FluoSpheres NeutrAvidin-Labeled Microspheres | Thermo Fisher | F8776 | |
BD Pharmingen PE Mouse Anti-Human CD45 | BD | 560975 | |
Bovine serum Albumin | MP Biomedicals | 8810025 | |
Corning V-bottom polystyrene 96-well plate | Corning | 3894 | |
Cytochalasin D | Sigma | 22144-77-0 | |
EZ-Link NHS-LC-LC-Biotin kit | Thermo Fisher | 21338 | |
Falcon 15mL Conical Centrifuge Tubes | Corning | 352196 | |
Falcon 50mL Conical Centrifuge Tubes | Corning | 352070 | |
Fixable Viability Stain 450 | BD | 562247 | |
Formaldehyde solution | Sigma | 252549 | |
HBSS without Calcium, Magnesium or Phenol Red | Life Technologies | 14175-095 | |
Human BD Fc Block | BD | 564219 | |
Human Serum Albumin | MP Biomedicals | 2191349 | |
Kimtech Science Kimwipes Delicate Task Wipers | Kimberly-Clark Professional | 34120 | |
PBS 1x pH 7.4 | Thermo Fisher | 10010023 | |
Polystyrene 10mL Serological Pipettes | Corning | 4488 | |
Protein Concentrators PES, 3K MWCO, 0.5 mL | Pierce | 88512 |